Вставка
- Опубліковано 14 жов 2024
- Visit Adobe Creative Cloud for more information: www.adobe.com/...
#VoCo is an audio manipulator that allows you to change words in a voiceover simply by typing new words. Presented live during the Adobe MAX 2016 Sneak Peeks, co-hosted by Jordan Peele. Learn more about this year's Sneak Peeks here: adobe.ly/2ffyder
Subscribe to Adobe Creative Cloud: www.youtube.co...
Download Adobe Express: adobe.ly/3rDM7M8
Learn More: www.adobe.com/...
About Adobe Creative Cloud:
Adobe Creative Cloud gives you the world's best creative apps so you can turn your brightest ideas into your greatest work across your desktop and mobile devices.
Connect with Adobe Creative Cloud:
Twitter: / creativecloud
LinkedIn: / adobe-creative-cloud
Facebook: / adobecreativecloud
Instagram: / adobecreativecloud
#VoCo. Adobe Audio Manipulator Sneak Peak with Jordan Peele | Adobe Creative Cloud
• 10X Your Excel Skills ... I3l4XLZ59iw
This is going to revolutionize RPG games. Finally the NPCs can refer to you by your actual character name and not some title pre-recorded.
Hurry, we don't have much time xXÍꟾꟾÍⱰåȵXx!
ROFL!
That would be great, but games already have this, i know that fifa does. There is your name, then your phonetic name is what it actually pronounces
So they can now afford to have any celebrity in any game with almost no work in their part. Or even more freakish, have games with the voices of people that are no longer alive. Spooky.
That's not how it works. Adobe also owns the rights to their software
Adobe: "we made sure you could never trust a photo online again, now were going to make sure you can't trust a sound bite"
"It's just words, it's just sound bites..." - Donald J Trump
and with face2face (www.graphics.stanford.edu/~niessner/thies2016face.html) you can never trust a video either.
I never trust videos after I learned after effects.. (but then I learned nuke and don't trust anything now)
i have a belly button
fukkkkkkk how am i suppped to enjoys the ethernets now¿
omg the memes are going to get to another level
It's gonna get so DANK!!
and porn
True
3d memes
This+deepfake
This is going to be really useful for people who make UA-cam Poops
I was thinking the same thing. Now I can't wait to make King Harkinian and Morshu say more bullshit
Oh my lord, I can see some sick Sentence Mixing right away.
same here, I gotta get my hands on this soon
That removes the skill of sentence mixing
AZMODIUS That’s a good thing because it makes YTPs more universally funny
getting a taped confession has never been easier
Anyone here after CDawgVA mentioned this on Trash Taste?
me!
i think this product is dead
same
@@CariagaXIII Example of something being too powerful to be allowed into the hands of the general population. The comments here show exactly why.
Which episode of trash taste is this? Have not been listening it lately.
So much for audio being admissible in court.
Audio has always been easily editable and manipulatable.
@@eddyavailable Not like this
Yeah, this will not be abused. Not at all :D
Time to make Obama say what he really means.
Or to make Trump sounds as if he's intelligent and can speak in complete sentences.
BobPagani Like Hillary? America great because America is good. wtf. rofl.
but first, it needs to be combined with very accurate lip-sync editing
You don't need lip-sync editing, the tech is already here: qz.com/654669/nothing-is-real-german-scientists-figured-out-a-way-to-make-putin-and-trump-say-anything/
Okay, the 'Jordan' edit sounded very fake, but the 'three times' was really impressive. Interested to see where this goes in a few years.
boy let me tell you
Great!!!! Now we can't ever believe Anything!!!!!
He mentioned watermarking audio around the ending, assuring viewers that there will be ways to detect if the audio is legitimate or not.
you could just cut it off in another program
Sam Flint What about watermarking the entire audio clip? The only thing I could think about is certain noise removal techniques, but if it's a specific kind of audio maybe it could bypass any noise removing techniques?
_"you could just cut it off in another program"_ I guess you don't understand how advanced watermarks are working today... There isn't a person saying "watermark, watermark" every 5 seconds. Todays watermarks are working with minimal audio transformations in a specific time. Like making it a verly little faster at 5 seconds for 0,4 seconds, putting also a low sound at 12 seconds. Even if you completly chance the audio like playing it at the PC and recording it with your mobile phone, then making it a very low quality a sound watermark would be detectable if you have the software key.
Also: Pictures can even today be manipulated or be staged, even video could be faked. People are aware of that and aren't believing everything they hear. And like he said: You need 20 minutes of sound for this, so only large speeches are possible to fake... And there are a lot more people who can witness that your recording is fake.
Nothing will chance in a negative way, but we could be profiting a lot from this.
Jan nis Thank you so much for explaining better than me, you worded everything so well and articulated in all the right ways n.n
Imagine a well-known voiceover artist recording a 30 minute sample of his voice that you can download for a fee, and from that point on you can use him for voiceover work saying whatever you want for your company or whatever. Basically you could buy someone's entire voice as a plugin for your projects. Crazy
This is crazy. I hope VoCo also gets released for other languages.
if it works the way i think it does and it is said to, it makes the new audio by analyzing the existing audio and -extrapolating- interpolating (so the sound isn't linked to a "how to pronounce English" database), the only thing that would change would be the voice to text part and the linking the characters to sounds, which we already have on our smartphones (i.e. siri can do voice to text) so I would assume its going to be released with other. This is just my guess though, i could be wrong.
Now I can embarass people more easily into Suicide......Seriously this can get really messy. Thats just one example.
can't wait to torrent this
hahahhaah
Charlie Indeed time to bring my Anime Waifu to life.
Charlie lolllll every Adobe product
Since it is exclusively cloud based, I would love if someone figured out how to bypass this! Better yet, I will join a class action lawsuit against them for collusion with Apple for making all their other software obsolete via software upgrades that is intentionally rendering previous versions useless and worthless
same
I expect a whole new era of UA-cam Poops.
Forward Synthesis micheal rosen will not be impressed.
I will be disappointed if this doesn't happen.
UA-cam poops are still alive and well, many popular YTPers are still making videos.
Even better when it comes to voice splicing. When people take several hundred clips and mix and match them so it sounds like they're saying completely different sentences. Currently, it sounds like several hundred clips were just mixed and matched - Voco could completely reshape that.
Or for video games. Imagine having your unique name said by NPCs or even inserting your own voice for your character.
Looking back, Jordan Peele REALLY saw the warning signs of this tech before the general public took notice.
The audience in this probably didn't think too much of it, especially since this program was never released onto the public and was quickly forgotten. But now, with AI Voice programs becoming widespread and for public use, we are really beginning to see the side effects of it. Sure, a lot of people will just use it for a quick laugh and for memes, but it's now become clear how kind of creepy this tech actually is, and especially when it gets into the wrong hands.
This can take player creation in sports games to a whole a new level.
Someday in the future it will be discovered that a famous podcaster had 1 year dead but his brother used this to keep his show online. That's an example of the kind of crazy stuff that will happen with these new technologies. It's so fucking exciting!
I was more thinking along the lines of "Yes I am Director X - please authorise payment of xxxx to account xxxxxxxx with sort code xx-xx-xx" having filler words pre-made etc. Lots of businesses now require an email for low amounts £5000~ maybe more.. doing this to 10 businesses a day would soon net a lot of money...
8 years ago this seems revolutionary. Now it’s quaint.
Some inspiration for Black Mirror season 4
Oh gawd, this is gonna be a game changer for angry political videos.
This is a perfect example of allowing technology to advance ahead of our ethical use of it.
Finally user/modded quests in The Elder Scrolls game can have fully voiced NPC dialog without actual voice acting! Sounds like something a synth would say...
Prank calls just got an upgrade!
yes!
OMG, I'm just blown away. Imagine the use in robotics. We struggle so hard with good text to speech.
You're a robot?
Josh B 😂😂
The Team Fortress 2 Source Filmmaker community will be all over this.
lol, HYPEE
This is so revolutionary! I don't understand why some people think this project is dangerous? Remember the early days of Photoshop, people were freaked out. But there's no big deal up until now, we can distinguish between original photo or a "Photoshoped" one. And it's actually changing photography industry so much. So just like Photoshop, I believe this VoCo project is gonna be a gold standard for future audio editing.
It doesn't require that far a stretch to think how this kind of technology can be potentially abused. Think resentful employees, prank callers, bullies and trolls, ruthless business competetors, falsifying evidence, kidnappers, spreading misinformation.
Think about it this way. What call could you receive on your phone right now that would ruin your day? From who and what would they say? Now, is it possible to create that with the software they're releasing?
This is mere speculation, of course. Not impossible things, however.
Poorly aged things
Finally, I don't have to work too much on my blackmail business.
Thomas Sijpkens Actually you'll have to work harder since people are less likely to believe in blackmain material now.
This is what the indie game industry needs.
One single game developer can make all the content for his or her game. . . except for the voice acting. Now all we need do is go on UA-cam or something and steal someones voice. Of course there is still the moral argument of if it''s ok to do that but on the other hand, alot of people sound like other people so it would be hard to make a claim that it's their voice they are hearing in the game as apposed to anyone else's.
My question is if it's possible to mess with and edit things like tone, speed and emotion.
As you might be able to generate the voice, in voice acting, you won't be able to generate the acting.
That's assuming that this tech doesn't support the things I questioned about.
Even still it will at least be able do stuff like when your character has a custom name. Normally you would have no voice, just text on screen. (exception, Fallout 4, which only had semi-custom names) With this kind of tech you can fill in the blanks.
You are definitely right about that! This opens up for a lot of cool possibilities.
I just hope they don't become scummy asshats and demand a subscription service.
People could give away their voice prints for free. For aspiring voice actors, could be a way of getting some exposure.
Wow amazing technology. Kudos to the developers. However I think we all know what all good technology ends up being used for.
Forged celebrity porno voice-overs...
aaaaand political/social/criminal damage to individuals
6:00
+Ramix09 Yeah, I watched the video.. I'm pretty sure effective watermarking is a given, but it's 2016, don't be so naive. The software can be reverse engineered, hacked etc. When billions and social control are involved, watermarking is just a minor setback. Peace...
+Lukos the only real use, finally it will have some decent VAs, even if they are computer generated /s
Be optimistic, not pessimistic.
Once upon a time people feared Internet, TV, Smart phone, even newspaper now they are part of our life. So be always optimistic.
This technology is going to usher in a new age of prank calling.
Yeeah.
To this day this is still the best voice synthesis I’ve ever heard
But they never released it. Now everybody else is catching up and nobody remembers Voco anymore.
Update me on your opinion now. 💀
@@knight808. the same 💀
"Oh, yeah. Oooh, ahhh, that's how it always starts. Then later there's running and um, screaming."
- dr Ian Malcom
"programmers"
So now we really can't believe what we see or hear..
Adam Udale I'm a audio engineer and switching words around is not hard..but to type in your words and mimic the voice is another form of forgery..
Adam Udale quite possible without Photoshop but not this easy..melodine will change a words timing ..but not easy to make new words..
and by what I see..you won't be able to tell with the wav form it's mimicking even the wav form..you gone fool anything voice recognition
There are people impersonating the voice of others since the existing of recorded sound. And this is way more real than a generated voice. No matter how good the software is. And people definitly never believed everything the hear. You can see it in the internet... People calling photos fake and "photoshopped" even if they are real. Even today nobody is believing anything without having multiple sources.
Jan nis Thats the point. Now who will tell us what is true? The media who has the money to "gatekeep" the truth or the common man blowing the whistle.
This is actually one of the best if not THE best sneak. I loved being there to see it for the first time.
Holy fuck.
Whaaaaat?! Very impressive! It doesn't sound fake or robot. Congratulations, Adobe!
One question: Will this just work for english speakers?
It is based on phon recognition/recreation, so no technical barrier for other languages.
See techcrunch.com/2016/11/03/adobes-project-voco-lets-you-edit-speech-as-easily-as-text/
The way we read depends on language.
So there is no chance to produce these results like on google translate?
translate.google.com/?hl=pt-PT#en/pt/Eu%20fui%20%C3%A0%20praia. (listen in english and in portuguese).
This will spawn an entire new genre of UGC. Well done, Adobe. Well done. :P
OK, I'll bite: UGC?
Richard Plunkett User Generated Content :)
***** Cheers.
Now we can change modern rap lyrics to make it comprehensible. Wonderful work.
so this is a product for the masses now.
make a guess since when Military agencies & FBI, CIA, NSA, etc. had access to such tech, and what they possibly did (were able to do) with it...
was my thought too
thought the same thing. also, i see copyright problem. people will use voices from celebs for Voice overs. maybe iam too negativ here, but it see no real worth for the consumer marked in this software other then abuse
eh, probably around 10-20 years ago.
sounds about right
sometimes the private sector develops something and then government agencies adapt it to their use, like xbox controllers for drone control.
I remember this, now look at what we have now.
"Believe nothing that you hear, and only half of what you see"
Now I will be able to understand Desiigner, Future & Young Thug?
Noldy HAHAHA
Noldy You're fucking dumb. It doesn't decode what someone has said it creates audio in their voice.
You're fucking dumb. You can't even decode a fucking joke on the internet.
ok, bye. Haha tell em.
no because it has to have actual speech. random mumbling doesnt count 😂😂😂😂😂😂
this is highly relevant to my interests.
Now the media will be able to change what someone says, bravo!
Also voice actors for games and animations/cartoons will now only have to do a few minutes work and then you can use this system to get infinite phrases in their voice. This either means that their wages are going to plummet to the floor due to less hours needed, or there will have to be some intellectual property law where you can passively earn income by selling the right to use "your" voice (more like a voice that sounds like yours technically) for how many hours or whatever. Whichever the case, this is going to completely change the labor market in voice acting.
Voice acting probably won't change much as a result of this; all it does is sentence mixing. It can't simulate emotion, and it sounds weird and stunted if you try to push it too far.
Dzvfars Fdszvfvf
True, but it seems to copy the same tone that the speaker was originally using, so in theory all you'd need is a sample of different emotional voice tones, and then you can say anything in those tones.
So you can have the voice actor scream painfully: "WHY DON'T YOU LOVE ME?"
And then the program can theoretically produce a sad scream of anything else. Yes, you'll still need lots of different samples, but the point is that you wouldn't need the voice actor to say all the lines, just all of the ways in which he or she can say anything else. You'd never need to read the whole script, just parts where you need to change tone or emotion.
You're gonna need to record more random emotions and random lines to get anything approaching something sounding natural in a project's voice acting. Any half-decent project will just find it easier to let the voice actor read the lines they need them to read. Lazy assholes will just record a few lines and make it sound half-way to Microsoft Sam, because lazy assholes never put much effort into hiring good voice actors anyway.
Well audio evidence is dead.
Yeah even with the watermarks probably.
You know you need 20 minutes of perfect recording without interruptions? I didn't know when I was talking that much without people speeking in the background the entiere time... And there are already really good impersonators... It is much easier to get them than using a software to fake speeches. And it would be much safer to use them than to use a digital generated sound.
Because even without a watermark it would be possible to detect that it is generated. There is no perfect way to imitate a voice perfectly and it never will. It always has errors you can hear if concentrating enough at it.
Jan nis
Fair enough.
May be in the future, they could build somekind of database or "sample Packs" to purchase separately, they would make more money, and have more reason to continue developing. So like changing they voice on a GPS, you change the voice from male to female etc. A voice database.
All you need then is 20minutes from that sampepack and you're good to go.
PRISM allowed the NSA to collect iOS, Google, and Microsoft data, including audio.
This is going to get a lot of people divorced hahaha
I don't think it will majorly impact courts. Records will just be treated similarly to photos. And by the way, you could have already mimicked someone's voice before, just much more slowly.
with this, we will never lose the voices of batman and the joker.
Imagine this being used in Voice Acting for video games! If you're creating DLC you can just expand on the voice acting with this and don't need the actors to come in as much. Double edged sword?
My question is if it's possible to mess with and edit things like tone, speed and emotion.
Or use two patters. So you can speak, the soft read some of the tones and mix with the pattern of other voice to make the first voice sounds like the second.
The tipical "voice copier" of fiction but real at least.
I think that if the pattern of a voice is smaller, perhaps even games can use this as a voice compressor, because store only the text, tone, speed and volume should be a very small info compare to a sound wave even with lossy compression like mp3.
Even some voice patterns could be altered or mixed by software between to create new patterns so a game could have a unique speak voice for each virtual character at a reasonable cost (only a small group of real speakers behind all scripting).
5 real speakes, 20 voice patterns, "infinite" software generated (by mixing) patters, to set a unique pattern for character.
You gotta remember, we are kind of behind in audio technology. All we can "create" before this was tones and pitches. Everything else (including this) is just manipulation of recorded audio that already exists. I can see why what you are theorizing can be a bit. . . difficult. I personally am just waiting for the day that we can paint with sound. Either way, this is definitely a step in the right direction.
I thought the exact same! Except more along the lines of, could games actually start to SAY our characters names?
Athedia Yes they can.
Easy to make this detectable. When words are added, an inaudible high pitch is layered on. Then in court cases (where the audio is fraudulent) they can detect the high pitch using the same software.
This could be invaluable for a current project I'm working on. I hope this comes out relatively soon. *crossing my fingers*
pls no human cyborg AI
Finally, we can get English dubs that sounds like the original voices!!!
The first part just sounds like something a child would do with Audacity.
The 2nd part though is truly amazing.
I doubt the software will be as good as this promotional video
Yeah I guess that was the strategy of the presentation to make people say oh so you just linking the audio with speech to text stuff. then it introduces all that learning stuff just like you expect from a presentation presented by an Asian geeky guy.
Havvse software might not be as good in real life applications, that's true. It was the case with Melodyne DNA at least, presentation demos were flawless but in real life a minor to major piano chord change is usually unusable.
for all we know, he said the words 'three times' at some other point in the speech and they pulled it from there. I'd like to hear it try to say something it can't just sample for, and with vocabulary the person never normally uses, like like tom cruise saying "dank memes can't melt steel dreams"
Calm down internet. It is supposedly using a large database of phrases and essentially does speechtext conversion by grabbing small phrases from a larger database, hence why this example is so limited. In the end it means we one can replace Siri with Morgan Freeman.
OH SNAP. THE IMPLICATIONS!
Holy shit. The impact for anything other than entertainment is staggering.
This is going to get horribly abused.
Exactly why I want it.
but the funny asian guy made it seem so innocent.....
how?!
You can ruin a person's reputation by creating an audio file of them saying something inappropriate, offensive, or illegal. You could take a foreign head of state's voice, create a fake speech calling for war and then use it for justifying military action. You could censor a person's press statement and make it more tame.
totally, and it'll destroy studio industry even more than it's already is.
This is streamlined sentence mixing, and I _love_ it! This is gonna be fun to mess around with!!
"There will be a watermark"... ah, we are now safe from malicious users... I can already think of three tricks for removing them, just by hearing the "There will be a watermark" sentence, LOL
watermark or not the way reality is changed with this has many implications. in the wrong hands this can be used to forgery and the only ones able to tell what is real and what is not are the few ones from Adobe that engineered this! nevertheless, right now they are trying to make it easier to detect! heck! that means the version we are seeing here is still too easy to be used for such purposes!
Someone is going to reverse engineer it, other techniques might be used to remove or smudge the audio watermark like converting to lower quality formats, filtering out frequencies that are inaudible to us (like mp3 does) that may contain said watermark. They'll figure something out. Unless it's made to be very obvious just from hearing it (without having to check a spectrogram or some similar process) that it is fabricated. This can potentially be used to do a lot of wrong, more than image editing can ever do.
ivvg sure thing! I was thinking something along those lines! this is like hacking reality! I can just imagine the things that this can cause should it fall in the wrong hands!
However it can also be used for good things. I think we should be able to hear the voices of dead voice actors saying things as if they were alive! that could allow some things in the entertainment environment that were impossible before!
I thought it was a lot cooler until the very end.. It seems like it just records a speaker for (currently around) 20 minutes, then the software picks out the individual words and collates them with the text version of the words the narrator was hired to speak? It's basically just a tool to more easily do what 'spies' do in movies where they get a person to 'say' all the words needed for a password, then on tape or digitally rearranging and cutting until its only the right words.
Basically, this wouldn't seem to make it exceptionally easier to fake 'your' voice unless it is extremely easy to mod their software and get 'your target' to specifically say and record the phrases and words you need. The real benefit is going to be in Adobe's growing voice library that is provided and how the tool makes integrating voice easier / real time. So it's less cool than I imagined, but actually a really good and useful idea/tool.
Oh shit. this is real.
God, I'm getting old. I used to edit this manually lol
if I listened to this without context I wouldn't have been able to guess that it was a text to speech program I would probably guess it was just bad audio quality. This is really cool, it doesn't sound robotic like over text to speech programs which is normally the biggest give away.
good news for prankers, criminals and journalists
For You Tubers like me with vocal issues, this will revolutionary.
this will be great for ytps
In my opinion, it will take the fun out of it. Part of the reason why sentence mixing in YTP is so fun is looking for the syllables and sounds to make new words.
Isn't that's what it is? Software recognizes voice and splits up words and syllables to form new voice patterns. It's YTPs automated.
Radi0he4d1 Yes, but you missed the point of what I was saying. What makes it fun is hunting for and trying to find the words, sounds, and syllables you need yourself.
I'mBored777
Exactly, and now the robot will do it.
This is NOT actually speech synthesis, it is only taking other parts of the speech that we don't hear and copying/cutting/pasting the words around. In other words, Keegan-Michael Key said the word "Jordan" and "three" and "times" somewhere else in the speech and the program allows you to edit the speech like text. That said, it is still a very useful tool. (As a disclaimer, I don't know if the program has the ability to synthesize speech or not, but that is not what is happening in this video).
Photoshop for voice over! congrats, Zeyu
I cannot believe these presenters' enthusiastic about it's application as a toy while not focussing on what this could mean in the wrong/everybody's hands! The possibility that this is already in use by certain factions or that it else would be developed by someone else does not change that at all. This asks for in-depth and open ethical considerations and discussions by the creators/vendors.
can we add emotions in the voice that we change or is it gonna have the same tone?
When... when will this be available?
they can record anything and can portray anybody's personality. this will get used in politics. this will be very harmful.
Generation "П"
A black mailer's dream
I suggest that everything made in this program should have a consistent frequency that humans can't hear to go with it so then we could have a computer analyze the audio to see if it has been altered with this software by listening for that certain frequency
Even then someone could find a way to remove that frequency and cover it up. Where there's a will, there's a way. And of course this goes both ways. Audio will no longer be trusted.
Oh that's true
Hey Adobe I Really Love Your SoftWares Its Really Helpful
so now you not only forge document, but voice recording too, nice.
And this is 3 years old. It's way beyond this now.
If this tech can be adapted in real time somehow, voice acting in dynamic games like RPGs is about to be revolutionised. Voice actors could say your characters name no matter what it is, as well as refer to specific quantities and other named elements.
On the flip side though, if this tech gets more convincing it's DEFINITELY open to abuse. All this talk of watermarking will be entirely null and void once hackers get a hold of this software. We've already seen how the media manipulated what politicians have said this year with a bit of video editing. They'd have a field day with this tech.
How about when they get over needing 20 minutes of speech? People will be framed by this tech. It has some scary implications for society.
yes
This is awesome !!
9/11 phone-calls on the plane... nobody could have faked those... oh wait.
Well of course Jews invented it and are coming for you name. They asked at the last meeting, who the fuck is this T-Bone fella?
15 years is a long fucking time for technological advancement...
Actually most technological interventions come within that same span of time to the consumer market there-after. (cellar phones, drones, wi-fi, digital photo manipulation, electric powered vehicles and etc).
The idea that the US gov is 15 years ahead of consumer products has been laughable for many years.
Laughable by who? Stop living under a wrong, drones were operable in the 90's.
People are freaking out over a crude yet flashy way of doing something they can currently do with free software like audacity and that's been possible since the dawn of audio editing. When I recorded an audiobook in my parents' basement back in 2015, and didn't feel like going back to do a second take, I just reworked the footage I already had and manipulated my voice to sound like the proper manuscript. If a T, K, or D sound was a little too soft, I'd replace it with another from a different patch of the recording. I once changed a he to a she just by popping in the hiss of an S and blending it with the H using a teeny tiny little crossfade. The result was seemless, and that was on Audacity, a whole year before this conference took place.
This audio clip in particular is so easy to work with. "My dogs" and "my wife" are both followed by a pause and have identical first syllables. Just find a nice little anchor point that both clips share (I'd aim for that explosion of sound when the M opens up to let the Y escape) then swap them out and blend them back together at the seems with a simple crossfade (or don't! The transition from M to Y is harsh enough that you probably won't even need one!) The result should be impossible to detect to the untrained ear. Certainly a lot better than this choppy speech to text to speech nightmare.
So yeah, this kind of thing can be abused, but don't blame adobe for this silly feature which I for one would never use. Blame Edison for the invention of sound recording.
Imagine singing songs with your fingers in the future
#Vocoartist
I'm in love with the voco - O.T. Genasis Xd
Mixey, you haven't heard of vocaloids have you? All ready a thing! search here for the Avana vocaloid its pretty good
This is gonna be horrible from a legal standpoint
You can use some softwares to see whether or not a, say, tape is audioshopped
until the signature is found and it can be removed.
And what happens if you play the audio over a phone call or the radio and the listener goes nuts and does something terrible?
exactly... Actually scary to think...
27POP27 WATERMARKS THO
Mans1ay3r needs to get on this.
Forget gamer poop, this could be used to restore cut content with no surviving dialogue. The Oblivion Breton male voice actor, Ralph Cosham, died a few years ago, so having this would be crazy.
Is it available now? It has been nearly 3 years. This would really benefit nervous speakers like me who want to put out great content but don't have the patient to make the audio presentable. But if I can just create a 20 mins perfect speech of myself then I can replicate it using this system to create great learning content.
Also please make the neural network architecture that is making this possible open source together with learned parameters.
what could possibly go wrong with this...
Tobias Richter Billy MAYS selling suicide putty.
Wow this is a hell of a break through.
The genesis of mass confusion...
After 2 years nothing was released, come on guys I need this software right now, money is not a problem!
Be ready for trump audio meme's
We don't need to edit or add speech for him, he's got himself covered in the meme department.
Trump : I made a VOCO for you and made you say : you have a chemical weapons.you have to pay or we will invade your country
🤣🤣🤣🤣🤣🤣🤣
Yep, NO LONGER CAN PRE-RECORD TAPES BE TRUSTED. Trump will say the N-Word before Mid-Terms no doubt now. LOL. I wonder how long the CIA, FBI and Omarosa have had this tech? Can no longer be trusted.
what practical use besides nefarious purposes is this gonna be used?
Dokes editing out cuts in audio recording
Imagine a promo video on FIVERR with Morgan freeman's voice over :D
While that is an awesome use for this tech, I imagine copyright would come into play. Is there not s a part about likeness? Wouldn't be surprised if Morgon Freeman was pissed off that someone used his voice without permission.
This is impressive! Reminds me of that scene in Sneakers: "My voice is my passport".
Could you imagine using this for ADR in a film. If the tech gets better then they wouldn't even need the actor to come in anymore, they could just use this program to add in the new dialog.
On shows with low budget we already do "frankenbites" to create words from phonemes - this just makes it way easier
***** Hell if this works really good then they will just need 20 minutes of talking then they could just throw this into an audio-book, or even a animated film script. I mean I wonder about the possibility of some kind of emotion being added.
Great. Because who needs those pesky human actors, right?
Aural Stimulation | Jeff Clement You still need 20 minutes of human actors, but now they could get something done faster. Though I would miss some of the great moments that happen when an actor does some improv. Also, I was talking about ADR. It would mean that we would be wasting less of the actors time and that they could go do more work.
Revenge Of Waldo I hear you, for sure. However ADR is a also source of revenue for actors, studios and post houses. Software like this could jeopardize that.
TNice tutorials comnt from you is legendary. Uncomplicated like your video!!!
Text to speech will no longer sound like R2 D2
but i like R2D2
When can we buy it ... 4 years now since that event
I really need this very much.l badly want to hear my dad's voice who is not in this world.l wish l could hear him
When will this be available for beta subscribers ?
So it's basically a better version of Adobe Audition?
Plausible deniability just got a lot easier
When does this go into the product?
Nothing is real anymore! D-:
Did you really type that???